Transform project ideas into production-ready codebases through intelligent validation, planning, and automated implementation with comprehensive demo mode for immediate usage.
- 🎭 Demo Mode: Immediate usage without API keys - try all features with fallback implementations
- 🔍 Idea Validation: Critical idea validation with external API integration and business viability scoring
- 🎯 Enhanced Planning: AI-powered feature breakdown into epics and implementable stories
- 🏗️ Smart Scaffolding: Language-aware project generation (Python/Node.js) with comprehensive templates
- 🔄 Micro-Prompt Queue: Test-driven development automation with multi-provider LLM support
- 📊 Experience Learning: Automatic failure analysis and guard-rail generation from development experiences
- 🐛 Issue Reporting: Automated GitHub issue creation with system diagnostics and log collection
- 🔧 Configuration Management: JSON schema validation with user-friendly setup
# Install from PyPI
pip install ai-idea-cli
# Or create a virtual environment first (recommended)
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
pip install ai-idea-cli
# Initialize configuration (enables demo mode by default)
idea init-config
# Validate an idea (demo mode)
idea validate "AI-powered task management CLI"
# Create a project (demo mode)
idea new "AI-powered task management CLI" --dry-run
✅ Success check: All commands work immediately with demo responses and guidance
# Edit configuration file
nano ~/.idea-cli/config.json
# Set your API keys and disable demo mode:
{
"demoMode": false,
"validator": {
"apiKeyEnv": "VALIDATOR_API_KEY"
},
"models": {
"apiKeyEnv": "ANTHROPIC_API_KEY"
}
}
# Set environment variables
export ANTHROPIC_API_KEY="your_claude_key"
export VALIDATOR_API_KEY="your_validator_key"
Command | Purpose | Demo Mode | Full Mode |
---|---|---|---|
idea init-config |
Initialize user configuration | ✅ Always works | ✅ Creates config file |
idea validate |
Validate project idea | ✅ Demo validation | ✅ Real API validation |
idea new |
Create scaffolded project | ✅ Basic template | ✅ AI-enhanced planning |
idea report-issue |
Report bugs automatically | ✅ Manual instructions | ✅ GitHub CLI integration |
Command | Purpose | Requirements |
---|---|---|
idea run-queue |
Execute implementation queue | LLM API key |
idea queue-status |
Show queue progress | Project with queue |
idea experience collect |
Log development failure | Project directory |
idea experience summarise |
Generate lessons from failures | LLM API key |
idea upgrade |
Apply lessons as guard-rails | Project directory |
Option | Purpose | Example |
---|---|---|
--dry-run |
Preview without execution | idea new "test" --dry-run |
--demo |
Force demo mode | idea new "test" --demo |
--skip-validation |
Bypass idea validation | idea new "test" --skip-validation |
--project-dir |
Specify project directory | idea run-queue --project-dir ./my-project |
Demo Mode (Default - No API keys required):
- ✅ Immediate usage for testing and exploration
- ✅ All CLI commands functional with fallback responses
- ✅ Basic project scaffolding with "hello world" templates
- ✅ Clear upgrade paths and configuration guidance
- ✅ Issue reporting with manual instructions
Full Mode (API keys configured):
- 🚀 Real idea validation with business scoring
- 🚀 AI-enhanced planning with detailed breakdowns
- 🚀 Complete TDD implementation cycle
- 🚀 Automatic lesson generation and guard-rail updates
- 🚀 Automated GitHub issue reporting
- JSON schema validation ensures configuration integrity
- User config stored in
~/.idea-cli/config.json
- Auto-detection of demo mode vs full mode based on API key availability
- Demo Mode: Provides neutral scoring with upgrade guidance
- Full Mode: Real validation using external validator API with business viability scoring
- Configurable score thresholds and acceptance criteria
- Demo Mode: Basic project structure with common patterns
- Full Mode: AI-powered breakdown into epics and implementable stories
- Technology stack recommendations based on idea analysis
- Demo Mode: Static templates with development guidance
- Full Mode: Micro-prompt queue drives TDD implementation
- Support for multiple LLM providers (Anthropic Claude, OpenAI GPT)
- Experience collection from development failures
- AI summarization into actionable lessons
- Auto-generated guard-rails from lessons
For Demo Mode (Default - No Setup Required):
idea init-config # Creates config with demo mode enabled
idea validate "test idea" # Works immediately
For Full Mode (API Keys Required):
# 1. Initialize configuration
idea init-config
# 2. Set environment variables
export ANTHROPIC_API_KEY="your_claude_api_key_here"
export VALIDATOR_API_KEY="your_validator_api_key_here" # Optional
# 3. Edit config to disable demo mode
nano ~/.idea-cli/config.json
# Change: "demoMode": false
The configuration file is automatically created when you run idea init-config
. Here's what each section does:
Demo Mode Configuration (Default):
{
"demoMode": true,
"retentionDays": 30,
"scaffold": {
"templateVersion": "main"
}
}
Full Mode Configuration:
{
"demoMode": false,
"validator": {
"endpoint": "https://api.example.com/validate",
"apiKeyEnv": "VALIDATOR_API_KEY",
"minScore": 0.7
},
"models": {
"plan": "claude-3-5-sonnet-20241022",
"queue": "claude-3-5-sonnet-20241022",
"summarise": "claude-3-5-sonnet-20241022",
"apiKeyEnv": "ANTHROPIC_API_KEY"
},
"scaffold": {
"templateVersion": "main"
},
"retentionDays": 30
}
Variable | Required | Purpose |
---|---|---|
ANTHROPIC_API_KEY |
Full mode | Claude API for planning, queue execution, summarization |
OPENAI_API_KEY |
Alternative | OpenAI API as alternative to Claude |
VALIDATOR_API_KEY |
Optional | Custom validator API for real idea validation |
Anthropic Claude (Recommended):
- Models:
claude-3-5-sonnet-20241022
,claude-3-haiku-20240307
- Get API key: https://console.anthropic.com/
- Install:
pip install anthropic
OpenAI GPT:
- Models:
gpt-4
,gpt-3.5-turbo
- Get API key: https://platform.openai.com/api-keys
- Install:
pip install openai
idea-cli/
├── idea/ # Core package
│ ├── cli.py # Main CLI with global error handling
│ ├── config.py # Configuration management with JSON schema
│ ├── demo.py # Demo mode implementations
│ ├── validator.py # Idea validation with demo support
│ ├── llm.py # Multi-provider LLM integration
│ ├── issue_reporter.py # GitHub issue reporting
│ ├── planner.py # AI-powered planning
│ ├── queue.py # Queue management
│ ├── runner.py # Story implementation engine
│ ├── experience.py # Learning system
│ └── scaffold/ # Project scaffolding
│ ├── python.py # Python project scaffolder
│ └── node.py # Node.js project scaffolder
├── config/ # Configuration schema and defaults
│ ├── schema.json # JSON schema for validation
│ └── default.json # Default configuration template
├── templates/ # Copier templates
├── .github/workflows/ # CI/CD pipelines
└── scripts/ # Development scripts
my-project/
├── .idea/ # Project metadata
│ ├── plan.json # Enhanced plan with epics/stories
│ ├── queue.json # Implementation queue
│ └── _logs/ # Development logs and failures
├── src/ # Source code
├── tests/ # Test files
├── lessons/ # Project-specific lessons
├── .github/workflows/ # CI/CD for the project
└── README.md # ADHD-friendly instructions
When errors occur, idea-cli can automatically create GitHub issues with:
- System information (OS, Python version, architecture)
- Configuration summary (sanitized, no secrets)
- Recent log files from
.idea/_logs/
- Steps to reproduce template
# Report an issue manually
idea report-issue "CLI command failed" --dry-run
# Automatic issue reporting on unexpected errors
# (triggered automatically with user confirmation)
If GitHub CLI is not available, idea-cli provides:
- Pre-formatted issue content
- System diagnostics
- Copy-paste ready templates
"Command 'idea' not found"
- Fix: Ensure package is installed:
pip install ai-idea-cli
- Check:
pip show ai-idea-cli
should show installation - Try: Restart your terminal or source your shell profile
"Configuration error: API key environment variable not configured"
- For Demo Mode: Run
idea init-config
and verify"demoMode": true
in config - For Full Mode: Set
export ANTHROPIC_API_KEY=your_key
and ensure"demoMode": false
"Failed to initialize config"
- Fix: Check permissions:
ls -la ~/
(should be writable) - Try:
mkdir -p ~/.idea-cli && idea init-config
- Alternative: Use
--config-dir
flag to specify different location
"Import errors for anthropic/openai packages"
- Fix:
pip install anthropic
(for Claude) orpip install openai
(for GPT) - Note: Only needed for full mode, demo mode works without these packages
Commands work but show demo responses
- This is normal! Demo mode is the default
- To upgrade: Set API keys and change
"demoMode": false
in config - Verify:
echo $ANTHROPIC_API_KEY
should show your key
- Use
--dry-run
for testing without API calls - Use
--demo
flag to force demo mode temporarily - Set appropriate model choices in config (haiku for speed, sonnet for quality)
# Enable verbose logging
export IDEA_CLI_DEBUG=1
idea validate "test idea"
# Check configuration
idea init-config # Shows current config path and status
We welcome contributions! Here's how to get started:
# Fork the repository on GitHub
git clone https://github.com/yourusername/ai-idea-cli.git
cd ai-idea-cli
# Install in development mode
pip install -e .
# Test the installation
idea --help
- Test demo mode: All commands should work without API keys
- Test full mode: Set API keys and test enhanced features
- Run quality checks:
ruff check . && ruff format --check .
- Test error handling: Verify issue reporting functionality
- Update documentation: Keep README and docstrings current
- Test both demo and full modes
- Include tests for new features
- Update documentation for user-facing changes
- Follow existing code style and patterns
MIT License - see LICENSE file for details.
🎭 Start in demo mode, upgrade when ready!